This study proposes to perform a clinical-metabolic-neuropathological correlation in dementia, in particular, primary degenerative dementia of the Alzheimer type (AD). We will use clinical data on behavior, mood, function, and cognition obtained in the year preceding death as markers of severity. Proton nuclear magnetic resonance spectroscopy (1/H MRS) will be used to analyze 6 brain areas obtained at autopsy from 75 Alzheimer's disease (AD), 25 controls, and 15 non-AD demented controls over 5 years. The first goal is to broaden the metabolic understanding of AD and to delineate clinical-metabolic-neuropathological correlations in a way that may provide insights into the timing of pathogenetic events over the course of this dementing illness. The second goal is to provide a detailed in vitro database for future extensions of this study into 1/H MRS studies of living patients with AD. No such detailed database currently exists. The metabolites measurable by 1/H MRS include N-acetyl- L-aspartate (NAA), L-glutamate, GABA, glutamine, myo-inositol, choline- containing compounds, creative and others. NAA is important because it is a putative neuronal marker easily detected by in vitro and in vivo 1/H MRS and can give an estimate of neuronal survival. Much like senile plaques and neurofibrillary tangles, NAA can be considered a new candidate marker of the neuropathological severity of dementia. The excitatory and inhibitory amino acids also play key roles in excitotoxic theories of several dementias. The choline-containing compounds include a phosphodiester which is a product of membrane degradation. In addition to determining differences between AD and control, demented non-AD brains will be examined to determine the specificity of the changes for AD. Clinical-metabolic and metabolic-neuropathologic correlations to NAA, senile plaques, and neurofibrillary tangles will be done in an attempt to determine which changes represent early, potentially causative, events and which changes are more likely secondary effects of neurodegeneration. In addition, a separately funded study will be analyzing the tissue by 31/P MRS and the levels of the membrane metabolites, phosphomonoesters and phosphodiesters, will be available for correlative studies as well. We hypothesize that markers of membrane proliferation and neuronal inhibition will be elevated early in the disease and decreased at later stages. In contrast, markers of membrane degeneration and excitotoxicity will be elevated at later stages. Preliminary results suggest that the in vitro 1/H MRS studies proposed in this application could provide information that is valuable in both a diagnostic and pathophysiologic sense and be readily extended to non-invasive, longitudinal studies of living patients which could aid in monitoring the course of the illness and tracking efficacy of experimental therapies.